34465A - rogue readings in DCV-HighZ mode

[Traceless]

Switching a KS 34465A to DCV/HighZ mode (with or without probes connected) causes the device to show an arbitrary rogue reading which steadily drifts further away from zero (not µV or mV far but multiple V far) Moving the hand close to the DMM or probes (without touching) resets the reading to an arbitrary different value and the process starts over. The manual does not give too many information about how HighZ works other that that it is supposed to be used for measurements that would be significantly influenced by the standard a 10 MOhm input impedance. Can anybody explain this effect? Since I've never owned a meter with >10MOhm input impedance I wonder how HighZ mode can be utilized properly given such high noise levels.

[HighVoltage]

That is very normal for the High-Z mode.

[bdunham7]

Can anybody explain this effect?

Input bias current.  The input amplifier leaks a small amount of current (10s of pA) into the front end and this charges up the small capacitance at the input at a steady rate.  So you'll see the numbers slowly climbing or falling as the case may be.  Every high-impedance voltage meter out there has this 'feature'.

[Kleinstein]

With open inputs the voltage reading in the high Z mode is pretty undefined. If at all a good external shield would be needed and than high noise and dirft is possible and expected. It is no rare that the meter even drift all the way to overflow.

With a reasonable low impedance signal source (e.g. < 1 M) the high Z mode should not behave much differente from the mode with 10 M input impedance, excepts for the missing loading of the source. The noise, drift and offset should be essentially the same. With the probes connected there should be a stablel zero reading (maybe a few µV off from thermal EMF at the probes, plugs).

This is not specific to the KS34465 but gneral the case for the DMMs with high Z ( > 10 G or similar).

[Traceless]

Thanks everyone for the explanations.

With the probes connected there should be a stablel zero reading (maybe a few µV off from thermal EMF at the probes, plugs).

I've just double-checked, even with the probes connected I've already drifted off almost 8V. That happens even with "Auto Zero" enabled.

[bdunham7]

I've just double-checked, even with the probes connected I've already drifted off almost 8V. That happens even with "Auto Zero" enabled.

He means with them shorted.  Just plugging them in doesn't do anything.

Just to be clear, the reason Auto Zero doesn't matter is that the 'stray' or 'phantom' voltage you are reading is in fact actually present at the input.  If you take two Hi-Z meters in parallel, they will drift off in unison due to their combined bias currents.

[Traceless]

Ah okay I misunderstood that. With the probes connected and shorted the meter reads ~ -3.5µV so all shiny in this case =)